Which states have AI chatbot laws?
An AI chatbot law, as used on this page, is a state statute that defines chatbots, companion chatbots or mental health chatbots and puts duties on the companies that operate them. As of November 11, 2026, fifteen states have at least one: California, Colorado, Connecticut, Georgia, Hawaii, Idaho, Iowa, Maine, Nebraska, New Hampshire, New York, Oregon, Rhode Island, Utah and Washington. California and New York have three each, which brings the total to 19 laws.
The list comes from the enacted section of the Future of Privacy Forum's 2026 Chatbot Legislation Tracker, which links each row to the state legislature's bill page, and from the effective dates in FPF's September 2026 report, Regulating the Conversation. FPF put the national count at 19 when New York's governor signed S 9051B in late October 2026 (FPF). Other trackers publish lower or higher totals because they count differently: some leave out disclosure-only laws such as Maine's, and some add licensing laws that restrict AI therapy without mentioning chatbots.
The laws arrived in two waves. FPF's report says five states acted in 2025 and that those first laws mostly required a non-human disclosure and crisis referrals. The 2026 laws go further, adding age assurance, parental controls, limits on engagement features and, in California, risk assessments and independent audits. This page sits under our state AI law tracker, which covers the wider body of state AI statutes.
AI chatbot laws by state: the table
Status is as of November 11, 2026. Effective dates are from FPF's Chart II unless a row says otherwise. The duty columns summarize FPF's coding of each law and are a guide, not a substitute for the bill text linked in each row.
| State | Law | Bill number | Status | Effective date | Disclosure duty | Minors rules | Therapy limits |
|---|---|---|---|---|---|---|---|
| California | Companion chatbots | SB 243 (2025) | In force | Jan. 1, 2026 | AI notice; reminders every 3 hours for minors | Content limits for known minors; crisis protocol | None coded |
| California | Adam's Law | SB 1119 (2026) | Enacted | Jan. 1, 2027; core child safety duties July 1, 2027; first audits by Jan. 1, 2029 | AI notice to minors | Age assurance, default parental controls, risk assessments, audits | Limits on diagnosing or treating health conditions |
| California | Companion chatbots in toys | SB 867 (2026) | Enacted | Jan. 1, 2027; repealed Jan. 1, 2031 | None coded | Access ban for toys with a companion chatbot | None coded |
| Colorado | Chatbot safety | HB 26-1263 | Partly in force | Aug. 12, 2026; most operator duties Jan. 1, 2027 | AI notice every 3 hours | Age assurance, parental tools, content limits | No claims of licensed health or legal services |
| Connecticut | Online safety act, AI companion provisions | SB 5 (2026) | Enacted | Jan. 1, 2027 | AI notice every 3 hours; hourly for minors | Parental tools, content limits | Mental health services only if designed for them; not a licensed professional notice |
| Georgia | Companion chatbots | SB 540 (2026) | Enacted | July 1, 2027 | AI notice every 3 hours; hourly for minors | Age assurance, parental tools, content limits | Therapy restriction coded |
| Hawaii | Companion chatbots | SB 3001 (2026) | In force | July 14, 2026 | AI notice; hourly reminders for minors | Parental tools, content limits | Therapy restriction coded |
| Idaho | Companion chatbots | SB 1297 (2026) | Enacted | July 1, 2027 | AI notice; every 3 hours for minors | Parental tools, content limits | Therapy restriction coded |
| Iowa | Companion chatbots | SF 2417 (2026) | Enacted | July 1, 2027 | AI notice every 3 hours | Parental tools, content limits | Therapy restriction coded |
| Maine | Chatbot disclosure | LD 1727 (2025) | In force | Sept. 24, 2025 | AI notice to all users | None; not minor specific | None coded |
| Nebraska | Conversational AI | LB 525 (2026) | Enacted | July 1, 2027 | AI notice; every 3 hours for minors | Parental tools, content limits | May not be programmed to claim it provides professional mental health care |
| New Hampshire | Child safety and generative chatbots | HB 143 (2025) | In force | Jan. 1, 2026 | None coded | Prohibited content for minors; harm response | None coded |
| New York | AI companion safeguards | S 3008C (2025 budget) | In force | Nov. 5, 2025 | AI notice every 3 hours | None; not minor specific | None coded |
| New York | Integrated AI companions | S 9008C (2026 budget) | Enacted | Jan. 1, 2027 | None coded | Age assurance, parental tools | None coded |
| New York | Unsafe AI companion features | S 9051B (2026) | Enacted Oct. 28, 2026 | Jan. 1, 2027 in the bill; FPF reports a chapter change to June 1, 2027 | May not present itself as human to minors | Age assurance; 12 categories of banned features | Therapy restriction coded |
| Oregon | Companion chatbots | SB 1546 (2026) | Enacted | Jan. 1, 2027 | AI notice; break reminders every 3 hours for minors | Content limits; engagement limits | None coded |
| Rhode Island | Companion chatbots | SB 2195 (2026) | Enacted | Jan. 1, 2027 | AI notice every 3 hours | None; not minor specific | None coded |
| Utah | Mental health chatbots | HB 452 (2025) | In force | May 7, 2025 | AI notice | None; not minor specific | Regulates mental health chatbots: disclosure, advertising and data sharing limits |
| Washington | Companion chatbots | HB 2225 (2026) | Enacted | Jan. 1, 2027 | AI notice every 3 hours; hourly for minors | Content limits; limits on sycophancy and engagement | None coded |
Two points about the New York rows. Holland & Knight's November 5, 2026 alert gives S 9051B an effective date of January 1, 2027 (Holland & Knight), while FPF reports that the governor signed a chaptered version that moves the date to June 1, 2027. We show both until the chapter text settles it. A fourth New York bill, S 9408 on AI companions in toys, passed the legislature, and FPF's tracker still listed it as awaiting the governor when we checked.
What do companion chatbot laws require?
A companion chatbot is an AI system built to sustain a relationship with a user, not to complete a single task. California's SB 243 defines it as a system capable of meeting a user's social needs through human-like interaction and a sustained relationship, as FPF summarizes it. Most 2026 laws use a similar definition and exempt customer service bots and tools limited to a narrow business purpose.
Across the laws in the table, four duties recur:
- Disclosure. Tell users they are talking to AI, and repeat the notice on a schedule, usually every three hours.
- Crisis protocol. Detect expressions of suicidal thoughts or self-harm and refer the user to crisis services such as the 988 Suicide and Crisis Lifeline.
- Minor safeguards. Block sexual content and content encouraging self-harm for minors, and in newer laws add parental controls or age assurance.
- Reporting. Publish or file reports on crisis referrals and safety protocols.
California and New York show where the second generation is heading. California's SB 1119 keeps a reasonable-measures standard but adds risk assessments, default settings that parents control, one-hour session and two-hour daily limits for children by default, and independent audits every two years. New York's S 9051B instead bans operators from giving minors any of twelve categories of unsafe features, including outputs that suggest the system is human, flattery, and use of personal information a minor shared more than twelve hours earlier (FPF). New York's definition of a covered AI companion reaches any generative system with a natural-language interface that gives ongoing, adaptive responses, so general-purpose assistants may be in scope (Holland & Knight).
These statutes followed lawsuits by families who allege that chatbots contributed to the deaths of teenagers, including Raine v. OpenAI. If you or someone you know is in crisis, call or text 988 in the United States.
Do chatbots have to say they are AI?
In the states with a chatbot law, usually yes, but the trigger varies. Maine's LD 1727 is a pure disclosure law that applies to all users. New York's 2025 law, Colorado, Iowa and Rhode Island require a notice at set intervals for every user, coded by FPF as every three hours. Connecticut, Georgia and Washington add a shorter interval, hourly, when the user is a minor. Hawaii sets a one-hour interval for minors.
California's SB 243 is narrower than it is often described: the general notice applies where a reasonable person could be misled into thinking the bot is human, and the three-hour reminder applies to users the operator knows are minors. Utah's HB 452 requires mental health chatbots to disclose that they are AI. New York's S 9051B approaches the issue from the other side, by making it unlawful to give a minor a system that presents itself as human, alive or capable of human emotions.
Outside these fifteen states there is no chatbot-specific duty to disclose, although general consumer protection law still prohibits deceptive practices. No federal statute requires chatbots to identify themselves as AI as of November 11, 2026.
Are AI companions legal for minors?
Yes. No state has banned minors from using AI companions outright. The closest measures are narrower: California's SB 867 restricts toys that contain a companion chatbot from January 1, 2027 until its repeal on January 1, 2031, and New York's S 9051B bars a list of companion features for minors without banning the products themselves.
What the laws do instead is condition access. Colorado, Georgia, California's SB 1119 and New York require some form of age assurance. Connecticut, Colorado, Georgia, Hawaii, Idaho, Iowa and Nebraska require parental tools or consent. California's SB 1119 lets parents disable access entirely for children under 16. New York's law defines a covered minor as a user the operator actually knows is a minor, and until state age assurance rules apply it lets operators rely on a user's self-declaration only when the user says they are a minor (Holland & Knight).
Enforcement differs as much as the duties. Most laws are enforced only by the state attorney general. California's laws, New Hampshire's HB 143, Oregon's SB 1546 and Washington's HB 2225 also let private plaintiffs sue, according to FPF's tracker. New York's S 9051B carries civil penalties of up to $25,000 per violation and allows a court to order destruction of unlawfully obtained data and of algorithms trained on it.
Which states ban AI therapy?
State limits on AI therapy come from two kinds of law, and trackers that mix them reach different counts.
The first kind is the chatbot laws above. Several bar a companion chatbot from holding itself out as a licensed professional or from offering mental health services except under set conditions. Connecticut allows an AI companion to offer mental health services only if it is designed for that purpose, its developers use clinical best practices and there are clear lines of accountability, and Nebraska prohibits programming a conversational AI service to state that it provides professional mental health care, according to FPF's report. Colorado extends the rule to legal professionals.
The second kind is professional licensing law. Nevada's AB 406 restricts offering or advertising AI-delivered therapy to the public unless a licensed professional conducts the service, FPF's report notes. Illinois adopted a comparable restriction in 2025. Becker's Hospital Review, citing a November 9, 2026 Telehealth.org report, counts seven states with bans: Illinois and Nevada from 2025, and Colorado, Maine, Rhode Island, Tennessee and Vermont in 2026. The same report says Delaware and Oregon bar nonhuman entities from using professional titles and that Utah requires disclosure instead of a ban.
These are bans on AI acting as the therapist. They generally do not stop licensed clinicians from using AI for administrative support. We have not verified each 2026 licensing statute against its bill text, so treat the count of seven as Becker's, not ours.
What federal chatbot bills are pending?
Congress has not passed a chatbot law. FPF's tracker lists three federal bills that have cleared a committee or a chamber as of our check:
- GUARD Act (S. 3062). Would bar AI companions for users under 18 and require age verification. The Senate Judiciary Committee approved it on April 30, 2026. See our GUARD Act page.
- CHATBOT Act (S. 4407). Sen. Ted Cruz's bill, built around family accounts and parental tools; approved by the Senate Commerce Committee.
- KIDS Act (H.R. 7757). Passed the House; its chatbot provisions require AI disclosure and limit content for minors. Our Kids Online Safety Act page covers the wider package.
The tracker also lists S. 4199, the Youth AI Privacy Act, as approved in committee. Whether a federal law would override the state laws in the table is unsettled; the Senate's GUARD Act text would leave more protective state laws in place. Our federal AI bill tracker follows each measure.
State bills still moving
Michigan's SB 760 and Pennsylvania's SB 1090 have each passed one chamber, and Arizona's HB 2311 was vetoed, according to FPF's tracker. FPF says it is tracking 98 chatbot-specific bills across 34 states, so the table above will grow in the 2027 sessions.
Fifteen states have enacted 19 chatbot-specific laws as of November 11, 2026, and twelve of those laws start in 2027. The common core is an AI notice, a suicide and self-harm protocol and safeguards for minors; California adds audits and New York bans listed features outright. No state bans minors from AI companions, and Congress has not passed a chatbot law.
What changed
- 2026-11-16: Page published, with status verified as of November 11, 2026.
- 2026-10-28: New York's governor signed S 9051B, banning unsafe AI companion features for minors.
- 2026-09-10: California's governor signed SB 1119 (Adam's Law).
- 2026-08-12: Colorado HB 26-1263 took effect; most operator duties begin January 1, 2027.
- 2026-07-14: Hawaii SB 3001 took effect.
- 2026-01-01: California SB 243 and New Hampshire HB 143 took effect.
- 2025-11-05: New York's first AI companion law took effect.
Frequently asked questions
Which states have AI chatbot laws?
As of November 11, 2026, fifteen: California, Colorado, Connecticut, Georgia, Hawaii, Idaho, Iowa, Maine, Nebraska, New Hampshire, New York, Oregon, Rhode Island, Utah and Washington. Together they have enacted 19 chatbot-specific laws. Six of those laws are fully in force and most of the rest begin in 2027.
Do chatbots have to say they are AI?
In most of the fifteen states, yes. Maine requires a notice to all users, and New York, Colorado, Iowa and Rhode Island require one at regular intervals. California requires it where a reasonable person could be misled and adds reminders for known minors. No federal law requires it.
Are AI companions legal for minors?
Yes. No state bans minors from AI companions outright. States instead require age assurance, parental controls and content limits. California restricts toys with a companion chatbot from 2027, and New York bars twelve categories of companion features for minors.
Which states ban AI therapy?
Illinois and Nevada restricted AI-delivered therapy in 2025. Becker's Hospital Review, citing Telehealth.org, reports that Colorado, Maine, Rhode Island, Tennessee and Vermont followed in 2026, for seven states in total. Several chatbot laws, including Connecticut's and Nebraska's, separately stop companions from posing as licensed mental health professionals.
Sources
- Future of Privacy Forum: 2026 Chatbot Legislation Tracker (links to each state bill page)
- Future of Privacy Forum: Regulating the Conversation, U.S. chatbot legislation report (Sept. 2026)
- Future of Privacy Forum: Back in the Chat, California and New York (Oct. 29, 2026)
- California Legislative Information: SB 243
- California Legislative Information: SB 1119
- New York State Senate: S 9051B
- Colorado General Assembly: HB 26-1263
- Holland & Knight: New York Enacts AI Chatbot Safeguards for Minors (Nov. 5, 2026)
- Becker's Hospital Review: The 7 states that ban AI therapy chatbots (Nov. 10, 2026)
- Congress.gov: S. 3062, GUARD Act
This page is journalism and commentary, not legal advice. AI law changes quickly; check the primary sources above and consult qualified counsel for your situation.
